SCIENTISTS trying to understand humanity's impact on climate are frustrated by not knowing what happens to climate naturally. Possible warming by carbon dioxide (CO2) pollution from burning coal, oil, and natural gas is a case in point.
Theory and computer simulations predict that a doubling of the concentration of this heat-trapping gas could warm our planet by several degrees Fahrenheit over the next half century. Some analyses suggest that this so-called CO2 ``greenhouse effect'' may already be evident. Climate data do reflect a warming trend. But it's different in character from what the computer models project. It may well be only a natural fluctuation.
In a comprehensive review of this problem, researchers at the Lawrence Livermore National Laboratory reached ``the clearly unsatisfactory conclusion that, while we are witnessing a warming of the terrestrial climate, we cannot identify its cause.''
They add: ``Even if it is of anthropogenic origin, it need not be due only to increased CO2. But if it's due predominately to CO2, then our present climate models require work to reconcile them with observational data on patterns of surface temperature change.''
The review team includes Hugh W. Ellsaesser, Michael C. MacCracken, John J. Walton, and Stanley L. Grotch. Their report appears in the current (November) issue of Reviews of Geophysics, a quarterly journal devoted to overviews of important research fields.
The climatic record is complex, patchy, and often tricky to interpret. Geological evidence, inferences from old agricultural records, and other non-instrumental data can be an unreliable guide. Even the instrumental records of the past century are marred by differences in observing standards in different regions and by changes of standards over time. Just moving a thermometer a few hundred feet can give a false indication of ``warming.''
Also, there's what the Livermore team calls ``the trap into which we fell because ... collection and archiving of worldwide weather records began in 1881, near the apparent [natural] temperature minimum of 1883.'' The team adds that, if information now available on that temperature trend had been known earlier, ``it is unlikely that there would now be the present degree of concern over the climatic effect of CO2.''